More > JRiver Media Center 29 for Windows

Too Easy VII -- Things we should do because they are so simple

<< < (11/21) > >>

Buldarged:

--- Quote from: Matt on July 19, 2022, 11:40:25 am ---Right now we switch from using the big thumbnails to the full size render at 400 pixels.  That's the size of the large thumbnail.

But the large thumbnail quality isn't the best so maybe we could try using 300 pixels next build.

Please test and let us know what you think.  Thanks.

--- End quote ---


Thank you for answering.
I do think I know what the problem is. The 400px cutoff value you are talking about is about the rendered width of a cover art in the view. At least that is what my tests confirm.
Edit: It is actually min(render with, render height). So the problem appears on landscape cover art as well.

However, as far as I understand the thumbnailing process, the larger side is always 400px. For series cover art that have a 2:3 aspect ratio that means the large thumbnails are of size 267 x 400.

So the view shows a cover art rendered at 400x600 using a thumbnail of size 267x400, which is why it looks blurry. If that is indeed the problem, the cutoff value should be left at 400, but compared to max(render width, render height) instead of just the render width. Of course this needs to be applied to all thumbnail levels. This would fix the problem for all aspect ratios at all size slider positions.

It might also be a good idea to the let the user decide the jpeg compression level of thumbnails. This setting would not affect old thumbnails, only new ones. Through the "Erase all thumbnails..." and "Build missing thumbnails..." options users could still update already created ones. Should also be implementable quite easily.
I personally don't care if I have 1 or 10GBs of thumbnails lying around, but other users might. Parametrizing allows for both. High quality thumbnails or low disk usage.

Even, if you only suceed in fixing the cutoff problem, that would still improve quality substantially.

DocCharky:
Translation bug in French.

Options / Video / "Display Settings"

In French, "display settings" is translated as "affichage des jaquettes" ("cover display") which make no sense and is probably the result of a mixup. I realized this when I spent 30 minutes looking for the framerate matching options... Correct translation would be "Paramètres d'affichage".

Matt:

--- Quote from: comox on July 01, 2022, 12:01:08 pm ---When I select a complex view that takes say 12 seconds to render, MC sometimes does not display the spiny thingy indicating it's thinking for the first 6 seconds of that period.

I know from experience that all is OK despite MC appearing dead, but it might be confusing for new users.

Good idea

--- End quote ---

Next build we'll try this:
NEW: A wait cursor is shown when filling list controls (for slow fills).

Testing appreciated.  Thanks.

Matt:

--- Quote from: DocCharky on July 21, 2022, 06:47:03 am ---Translation bug in French.

Options / Video / "Display Settings"

In French, "display settings" is translated as "affichage des jaquettes" ("cover display") which make no sense and is probably the result of a mixup. I realized this when I spent 30 minutes looking for the framerate matching options... Correct translation would be "Paramètres d'affichage".

--- End quote ---

Please feel free to edit the language file however you think best.  Then submit the edits to matt at jriver dot com for everyone to enjoy.  Thanks.

markf2748:
Left Panel > Services & Plug-ins > Media Network > Time column.
Change time display from hh:mm to hh:mm:ss.xxx, i.e. simply (?) re-format the display to include seconds and milliseconds.
 
Current display is frustratingly coarse in time.  The goal is to make it easier to compare MC report with WireShark analysis, and generally to get a more useful, finer-grained look at the sequence of events between MC servers.

Reference:  Primarily Item (1), along with some related things to consider:
https://yabb.jriver.com/interact/index.php/topic,130925.msg907639/topicseen.html#msg907639

Navigation

[0] Message Index

[#] Next page

[*] Previous page